Portrait of Kartik Khosa

Kartik Khosa

I’m a software engineer at Microsoft Core AI, where I build the agentic workflows that automate .NET migrations in GitHub Copilot. Before that I spent three and a half years modernizing The Arabidopsis Information Resource and MorphoBank, and I still build open-source tools for evolutionary biology.

Email me LinkedIn Google Scholar GitHub

My career as a cladogram, drawn the way OpenClade draws trees.

Experience

  1. Apr 2026 – now Current

    Software engineer, Microsoft Core AI

    Redmond, Washington

    • Agentic .NET migration engine. Engineered hybrid C# agentic workflows powering the official GitHub Copilot app, orchestrating Roslyn static code analysis and LLM synthesis across a three-stage engine (assessment, planning, execution) to automate end-to-end .NET Framework migrations, reducing manual effort by 90%.
    • Enterprise modernization at scale. Partnered with .NET experts to build modular upgrade plugins, accelerating multi-year first-party .NET Core migrations, saving years of developer effort and on track to eliminate millions of continuous CPU cores.
    • Autonomous bug triage agent. Developed an autonomous bug triage agent for Azure Storage Explorer using a stateful directed graph loop: GitHub issue ingestion, reproduction script synthesis, defect localization, unit test generation and pull request delivery with root-cause provenance. It resolved 85% of issues.
    • Industry leadership. This work contributes to Microsoft’s position as a Leader in the Gartner Magic Quadrant for AI-Augmented Code Modernization Tools (August 2026).

    C#, .NET, Roslyn, GitHub Copilot, agentic AI, LLMs, Azure

  2. Oct 2022 – Apr 2026

    Full stack software engineer, Phoenix Bioinformatics

    The Arabidopsis Information Resource (TAIR) and MorphoBank

    • Search architecture. Architected a Spring Boot pipeline replacing SQL queries with S3 JSON data indexed in Solr, cutting search times from 60 seconds to 300 milliseconds.
    • AI summaries. Led development of a LangGraph agent architecture that generates bioinformatics summaries for more than 50,000 locus pages, with section-based summarization, combination, validation and evaluation layers, powered by Gemini 2.5 Flash.
    • Automation. Automated reporting workflows with Spring Boot APIs, dynamic data aggregation and AWS SES, saving business teams more than 1,000 hours a year.
    • DevOps. Built a Docker-based CI/CD platform (Jenkins and GitHub Actions), speeding up releases and greatly reducing rollbacks.
    • Microservices. Designed RESTful Node.js microservices to modernize the Arabidopsis.org backend.
    • Research. Co-authored peer-reviewed papers in Genetics and presented at ISMB 2024 on scalable bioinformatics architecture.

    Java, Spring Boot, Vue.js, React, Node.js, AWS, S3, Solr, MySQL, Docker, Jenkins, Supabase, OpenAI

  3. Feb 2020 – Oct 2022

    Solution engineer, Appian Corporation

    • Security and scale. Managed authentication protocols (SAML, OAuth, LDAP) for more than 100 businesses serving over 100,000 users.
    • Code quality. Assessed and optimized more than 25 custom Java plugins against company standards.
    • Documentation. Wrote internal documentation that cut resolution times by 80% across more than 10 customer cases.
    • Client success. Achieved a 95% “very satisfied” rating from clients through technical consultation.

    Java, SAML, OAuth, LDAP, Git, technical consulting

Research

  1. 2026

    Recent major changes to TAIR: updates to the database, website, and Arabidopsis genome

    Genetics, Oxford University Press

    Database enhancements, website improvements and genome annotation updates to The Arabidopsis Information Resource.

  2. 2024

    The Arabidopsis Information Resource in 2024

    Genetics, Oxford University Press

    Our work modernizing TAIR to make it more accessible and useful for the global plant research community.

    Top 25% of research outputs; 87th percentile for its age, 84th overall.

  3. 2024

    Building a future-proof digital resource: the comprehensive modernization of TAIR

    Talk at ISMB 2024

    Presented at ISMB, the largest international conference in bioinformatics and computational biology.

  4. 2023

    Unlocking endless possibilities with GPT-4

    HackerNoon

    From study plans to building several applications with GPT-4. Trended at #4 for more than two weeks.

  5. 2019

    Framework for uncertainty quantification of a chemical kinetic parameter in MOCVD

    Honors thesis, Penn State

    Improved simulation efficiency by 85% using orthogonal sampling and stepwise regression, and determined yield ranges of nanomaterials through sensitivity analysis and Monte Carlo simulation.

Projects

  1. 2026

    OpenClade

    Phylogenetic analysis of morphological data in the browser

    Parsimony analysis of morphological character matrices that runs entirely in the browser, with no installation and no data upload. It reads NEXUS and TNT files, including MorphoBank exports, handles inapplicable characters with the Brazeau, Guillerme and Smith (2019) algorithm, and produces most parsimonious trees, consensus trees, bootstrap support, character mapping and a methods paragraph. One click shows which clades depend on how inapplicable characters are treated.

    Tree lengths match TreeSearch on all 806 test trees across 31 published matrices, and the search found the exact optimum in 124 of 124 exhaustive tests.

    Open OpenClade User guide Source code DOI 10.5281/zenodo.23634128

Press

  1. 2026

    How technical? The debate about AI’s future jobs

    Forbes, quoted

    Quoted as a Microsoft Core AI engineer on forward-deployed versus AI engineers, and on turning AI capabilities into measurable business value.

  2. 2024

    Running AI on device

    Cybernews, quoted

    Expert comment on how effective small AI models are on consumer devices.

  3. Jan 2024

    Leading the charge in modernizing biological data systems

    Free Press Journal, feature

    On modernizing Arabidopsis.org, including data pipeline work that cut retrieval times from over two seconds to under 100 milliseconds.

  4. Oct 2023

    Technology leadership profile

    DNA India, feature

    A profile of my path and contributions in technology.

Judging and review

  1. 2025

    Reviewer, BOSC 2025

    Bioinformatics Open Source Conference, ISMB 2025

    Technical review of submissions on open-source bioinformatics tools and computational methods.

  2. 2024

    Reviewer, BOSC 2024

    Bioinformatics Open Source Conference, ISMB 2024

    Technical review focused on open-source bioinformatics tools and AI/ML approaches.

  3. 2024

    Judge, SIIA CODiE Awards

    Education technology categories

    Assessed innovation and impact in educational software products.

  4. 2024

    Judge, Artificial Intelligence Excellence Awards

    Business Intelligence Group

    Evaluated AI business innovations, with a focus on sustainable and socially beneficial practice.

  5. 2023–24

    Committee member, Community and Travel Grants

    US Research Software Engineer Association

    Reviewed grant applications that support diversity, inclusion and community in research software engineering.

  6. 2023

    Lead judge, Information Technology World Awards

    Globee Awards, 18th annual

    Reviewed more than 100 submissions from technology leaders and products worldwide.

Education

  1. May 2024

    Master’s in Computer and Information Technology

    University of Pennsylvania

    Coursework in big data analytics, machine learning, databases, blockchain and cryptography, computer systems programming, algorithms, and data structures and software design.

  2. Dec 2019

    Honors B.S. in Mechanical Engineering, magna cum laude

    Pennsylvania State University

    Honors thesis on uncertainty quantification in metal-organic chemical vapor deposition (listed under Research).

Skills

AI and machine learning
Agentic AI (LangGraph, GitHub Copilot), LLM synthesis and prompt engineering, OpenAI and Gemini APIs, PyTorch, TensorFlow and Keras, scikit-learn, NLTK
Backend
C# (.NET, Roslyn), Java (Spring Boot), Python (Django, Flask, FastAPI), Node.js, C++, REST APIs, microservices
Frontend
JavaScript, Vue.js, React, Next.js, HTML and CSS, Tailwind CSS, Bootstrap
Data
MySQL, MongoDB, Redis, Solr, Supabase, Amazon RDS and S3, PySpark, Databricks, data mesh architecture
DevOps and cloud
AWS (EC2, Lambda, SES), Azure, Docker, Kubernetes, Jenkins, GitHub Actions, CI/CD, Bash

About

I live in Bellevue, Washington. Outside work I never miss a Formula 1 race weekend, I go skydiving and scuba diving when I can, I play Valorant competitively, and I’m always looking for a new place to eat.

I try to bring the same curiosity I have for those things to my work, whether I’m optimizing a system, improving an interface or building something new.

Illustration of Kartik at a café table with a coffee and a phone

Contact

I’m happy to talk about agentic AI, .NET modernization and research collaboration. Email is the fastest way to reach me.